IMPROVING MY PREDICTIVE MODEL ...
First of all, I worked on my betting model. The past model started to collapse in March. Something was definitely not OK, and I was forced to search for possible mistakes and ways to improve the model.
After two months and processing a mass of data (from all possible perspectives :)), I can't say with certainty that I found THE exact mistakes. I am not even sure if it's possible to state that, when you handle with so many random factors like in this case.
Foundations of the previous model seemed OK, and I think it's not advisable and even possible to come up with something completely new. I had a feeling that it's all about just small improvements, which of course, can have a huge impact on final results. I focused on selectivity, on decreasing the number of predictions and on decreasing the variance (volatility)
Finally, I added one significant factor and implemented few exclusion conditions to limit the number of predictions. In the late August and in September I have been doing back-test (for two years back) - that was a pain in an ass ... but the results I received turned out to be pretty good, so ...here I am again :)
FREE PICKS ...
I have a dilemma with this one... One the one hand, free picks is such a classic but on the other hand, I find it a kind of anachronism ... There are so many better ways to publish prediction (bets) and keep in touch with the followers nowadays.
For me still, THE MAIN REASON for this blog is to share my passion of sports predictions and of course to draw your attention to my paid service, which I truly believe is worth of attention because I know the effort standing behind it.
For those who want to follow my work, the best way, for now, is Omnibet (for paid predictions) and Blogabet (for all predictions). So for free picks (from time to time), you can just follow me on blogabet - I am sure that the notifications there are more efficient than the notifications via widgets available on Blogspot...
WHAT INSTEAD OF ...FREE PICKS
..but instead of free picks, I have come up with sth new. I hope this is something that you will like and I will have more fun with.
Let me introduce to you "GO DEEP/GO DOWN list" :))
What it is? In short, it is a sort of outright prediction (outright bet). Once a week before the tournament starts I will publish a list of players that I predict can GO DEEP in the tournament or will GO DOWN quickly. And then the fun starts... ;) As the week will go by I will follow the picked players and comment on their matches/results. What is new and I guess unique, (for I have never seen sth similar on other blogs) that by "going deep or down" I mean not necessarily winning but, as a handicapper, I mean covering the spread. Basically, I will analyze the line-up with regard to players' form, the draw, results in previous years, court preferences, current fitness, health issues, motivation etc. and according to that I will pick around 3 players for each list. Covering the spread is what counts in this game:). For example - player form "go down" list can even reach the semi-final but if she will cover the spread, let's say, in 1 of 4 matches then it will be considered as a success.
Likewise, if a player from "go deep" list will be out even in the second round but she will cover the spread in both of these rounds then it will also be considered as a success. I think if you retrace posts from one or two tournaments then everything gets clear for you. Enjoy!!
0 comments:
Post a Comment